i
ALUMNI IDENTITY: A STRUCTURAL EQUATION APPROACH TO BELIEFS AND BEHAVIORS
by
MICHAEL MCNAMEE III B.A., University of Wyoming, 2010 M.P.A., University of Wyoming, 2015
A dissertation submitted to the Graduate Faculty of the University of Colorado Colorado Springs
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Department of Leadership, Research, & Foundations 2020
© 2020
MICHAEL MCNAMEE III ALL RIGHTS RESERVED
ii
This dissertation for the Doctor of Philosophy degree by Michael McNamee III
has been approved for the
Department of Leadership, Research, and Foundations by
Sylvia Mendez, Chair
Megan Bell
Dick Carpenter
Phillip Morris
Kevin Rask
iii
McNamee, Michael III (Ph.D., Educational Leadership, Research, & Policy) Alumni Identity: A Structural Equation Approach to Beliefs and Behaviors Dissertation directed by Associate Professor Sylvia Mendez.
ABSTRACT
The relationships between alumni identity, engagement, and donations have been widely studied, but their true nature is still largely unknown due to the diverse factors suspected to contribute to donor motivations. Using the framework of organizational identity (OID) theory, this study comprehensively examined the way in which these constructs interact with each other using structural equation modeling (SEM). The alumni community of a small, selective, liberal arts college served as the population of interest. The data consisted of an alumni survey, institutional engagement information, donation records, and demographic information. The three tested models differed in how donations were measured and utilized separate outcome variables: donor status, cumulative giving, and number of gifts. The results suggested different relationship characteristics in each model, emphasizing the importance of using multiple metrics in alumni donation
research. Engagement was found to significantly predict both OID and donations in every model, and its significance highlights OID theory’s deficiency in accounting for relevant post-graduation experiences. The donor status model most supported OID theory by showing the importance of identity in initial support behaviors. The other models did not support the theory and emphasized the importance of using engagement as a predictor variable. Engagement was a stronger predictor than OID for both total giving and the number of gifts. While engagement is a consistently strong predictor of donations, there is evidence of the relationship being mediated by OID. OID was not significant in
iv
predicting total giving. The models advance the understanding of the field by showing that while both engagement and identity are key to predicting donations, identity is more crucial to the initial giving decision, and engagement is key to formation of long-term giving habits and increased amounts of support.
v
DEDICATION
This dissertation is dedicated to Hannah. Marrying her was the best decision I ever made, and she has held me up during this whole process while showing
vi
ACKNOWLEDGMENTS
It took a lot of people to get me to this point, and I am blessed with many to thank. First, I am eternally grateful for the members of my committee: Dr. Sylvia Mendez, Chair; Dr. Dick Carpenter, Methodologist; Dr. Kevin Rask; Dr. Megan Bell; and Dr. Phillip Morris. They have provided me with continual encouragement and the support I needed to stay on track. All the committee members have been generous and prompt with feedback, and they provided key ideas that strengthened the study.
Since I was working full-time during this process, I am grateful to my leaders at the college: Anita Pariseau, Sean Pieri, Mark Hille, Kristin Love, Dr. Jill Tiefenthaler, and Dr. Mike Edmonds. They not only provided the flexibility to attend class and conduct research, but also gave me numerous professional opportunities, plenty of pep talks, and opportunities to incorporate my research into my practice.
When I embarked on obtaining a PhD, I sought out leaders in higher education as role models. Through many coffee chats and phone calls, I gained more from their words than any of the books I read. While all these guides are very important to me, those of special note are Dr. Homer Wesley, Dr. Sara Axelson, Dr. Lance Bolton, Dr. Meghan Stidd, Dr. Jay Dillon, Dr. Lori Ann Thompson, Keener Fry, Dr. Nancy Hernandez and Dr. Noah Drezner.
My family has been incredibly supportive, and I am grateful to my dedicated proofreaders, Mom and Dad. My sisters and extended family have been incredible, and I owe special recognition to my cousins and soon to be PhDs, Ty and Chase McNamee.
Finally, I could not have accomplished this without the dogged support of my best friend, Kevin.
vii
TABLE OF CONTENTS CHAPTER
I. INTRODUCTION……….………..……..….. 1
Statement of the Problem...………..…….….……2
Purpose of the Study……..……….…………..……….2
Research Questions……..……….……….………...…….3
Theoretical Framework……….……….……..……..4
Significance of the Study……….……….….……9
Definition of Terms………..………...……….……10
Summary………..…………...………12
II. LITERATURE REVIEW………..….…………..……...……...…...13
Restatement of the Problem……….……….13
Overview……….………….………..……...13
Historical Background………….….……….………..…….14
Relevant Theories……….………….………..…….15
Altruism-Based Philanthropy Models ………..…….16
Exchange Theories………….………….………...…17
Role Identity Theory……….………….………19
Identity-Based Motivations.………….………..21
Significant Donation Predictors……….………….…………...22
Fundraising Practices ….……….….………..………23
Economic Factors….……….………..………23
viii Age……….……...……….………..………25 Racial Identity……...….……...……….………..………25 Degree Type………..….……...……….………..………26 Student Experience………..….……...……….………..…..………26 Alumni Engagement..….……...……….………..………27
Structural Equation Modeling (SEM)...….……...……….………27
Summary……….………….………..28
III. METHODOLOGY………..……30
Restatement of the Purpose………...….30
Research Design…………...……….………...30
Research Setting………..………….………...………...……31
Population and Sample………..33
Instrumentation………..35 Alumni Survey………...35 Alumni Profiles………..36 SEM Variables………...…38 Missing Data....……….……….40 Data Analysis……….……41 Approach………41
Path Analysis Diagram………..42
Control Variable Analysis………..44
Limitations……….45
ix
SEM Model………...………….…...….50
Number of Gifts Model………..50
Total Giving Model…..………..52
Donor Status Model…….………..54
Model Comparisons….………..57
Summary………...………….…...….58
V. DISCUSSION AND CONCLUSIONS..………..60
Summary of the Study...………...………….…...….60
SEM Application...………...………….…...….62
Research Question 1..………...………….…...….65
Donor Status Model…….………..67
Total Giving Model…..………..70
Number of Gifts Model………..73
Research Question 2..………...………….…...….75
Implications and Recommendations..………...………….…...….79
Future Research Directions………....………...………….…...….84
Conclusion……….………....………...………….…...….88
REFERENCES………..91
APPENDICES A. Selected Survey Instrument Items and Results………97
x
LIST OF TABLES TABLE
1. Location, Gender, Graduation Years, and Donor Status of the Survey
Respondents.………..34
2. Variables and Measures……….39
3. Survey Item Descriptive Statistics……….48
4. Non-Survey Item Descriptive Statistics……….49
5. Number of Gifts: Path Coefficients between Latent Variables……….52
6. Total Giving: Path Coefficients between Latent Variables………...53
7. Donor Status: Path Coefficients between Latent Variables……….…..56
xi
LIST OF FIGURES FIGURE
1. Proposed Correlates of Organization Identification ...………..6
2. Path Diagram Path Diagram of Engagement, Alumni Identity, and Donations...44
3. Modified Path Model for Number of Gifts………51
4. Modified Path Model for Total Giving………..54
5. Modified Path Model for Donor Status……….56
1 CHAPTER I INTRODUCTION
Fundraising is not only crucial for the modern-day operations of colleges and universities, it has been an integral part of the higher education movement in America since its beginning. Even the United States’ first institution, Harvard University, was launched into sustained prominence by its namesake, John Harvard, who generously donated an estate gift and a large book collection shortly after the college’s charter (Hood, 1991). That spirit of philanthropy has remained present in higher education, especially with the reliance of alumni support. According to the Council for Aid to Education (CAE), approximately 10% of college and university expenditures each year are directly covered by philanthropic support (Seltzer, 2018). This reliance on donations is even more present at private colleges, where the institutions are not bolstered by direct government funding. At these colleges, a steady endowment is a safety net against temperamental admission fluctuations and an expensive competition to demonstrate excellence in the areas of educational quality, facilities, and amenities (Horn, 2018).
Additionally, the situation is growing increasingly dire as high-profile college closures at Southern Vermont College, Green Mountain College, Marygrove College, and Newbury College serve as cautionary tales about what happens when financial instability occurs (Fain, 2019). Demographic trends bring even more dire circumstances for colleges relying primarily on tuition. A significant drop in birth rates occurred around the Great Recession, and there simply will not be enough incoming traditional students to fill current college campuses in the coming years. Horn (2018) predicted that 25% of institutions will close or go bankrupt in the next 10-15 years because their financial
2
models are not prepared for this downturn amidst growing competition in new forms of education. The only way to guarantee long-term success for private colleges is to shift away from tuition reliance and build large bases of philanthropic support, primarily from those who care most about the colleges’ success, their graduates.
Statement of the Problem
To accomplish long-term alumni support, it is imperative for colleges to cultivate and maintain high quality alumni affinity, and for alumni to see themselves as active members of the college community long after graduation. The problem with
accomplishing this goal is that practitioners in institutional advancement operations do not have a complete understanding of the components that go into building a relationship and leveraging the relationship for support. The field of predicting affinity and donor behavior is well studied, with significant advances in donor prediction models based on demographics and collegiate experiences (Drezner & Huehls, 2014). However, no known studies have examined alumni identity characteristics through a comprehensive structural equation model (SEM), taking into account the emerging idea that the construct of alumni engagement both influences, and is influenced by, alumni beliefs about their connection to the college. By doing so, this study advances the field by comprehensively explaining the relationships between four important factors: engagement, alumni identity, donations, and demographics.
Purpose of the Study
The purpose of this study is to visualize and determine the relationship between alumni identification with the college, their engagement with it, and their donation behaviors. SEM was used to understand these relationships because of the method’s
3
versatility and usefulness in mapping complex connections. The proposed model tested the idea that engagement affects identity, which then affects donations. It also tested if the various measures of donations change the relationship characteristics. This
methodology has never been applied to the study of these constructs’ relationships. This study used Mael and Ashforth’s (1992) organizational identity (OID) theory as its framework, which hypothesizes that alumni support their college due to a sense of connectedness stemming from key antecedents. While the study used OID framework, it departed from the original model by using alumni engagement as a predictor of donation behaviors, specifically testing how engagement influences, and is influenced by, alumni identity. The research was conducted on a sample of the alumni population of a small, selective liberal arts college in the western region of the United States. The college was particularly ideal for this research due to the presence of the OID college antecedents of distinctiveness, prestige, intercollege competition, and low intra-college competition.
Research Questions
This study sought to gain a better understanding of alumni donor motivations and how beliefs and behaviors serve as philanthropic predictors. Specifically, the following research questions guided the study:
1. What is the nature of the relationships between engagement behaviors, alumni identity, and donations?
2. Do the relationships follow Mael and Ashforth’s (1992) OID framework? To answer these questions, the study used data from an alumni survey and college records from a selective liberal arts college. Twelve survey items were synthesized into a construct of alumni identity. Similarly, an alumni engagement construct was derived
4
from four engagement activities that have suspected relationships to OID: event
attendance, volunteering, digital engagement, and being the parent of a current or former student. Six additional variables were used as controls in the SEM: estimated household income, class year, race/ethnicity, marital status, degree concentration, and geographic distance from the college. These combined measures were used in the model to predict donations on a representative sample of 1,922 alumni survey respondents. Donations were measured in three distinct ways: donor status, total number of lifetime gifts (compared to peers), and total lifetime donation amounts (compared to peers).
Both research questions were answered by the SEM. The first is answered by mapping out the various pathways and determining their strength and significance. The results are valuable in adding to the understanding of the way in which giving behaviors can be framed and predicted. The second question was answered by comparing the results of the analysis to Mael and Ashforth’s (1992) proposed framework.
Theoretical Framework
This study used OID theory to frame the reasons alumni choose to donate money to their alma mater. This theory is prominent in the field because it recognizes alumni often view the college as a personified entity, and they can strongly identify with it. According to Mael and Ashforth (1992), OID is “a perceived oneness with an organization and the experience of the organization’s successes and failures as one’s own” (p. 103). Mael and Ashforth derived this theory as a convergence of organizational management principles in combination with Turner and Reynolds’ (2010) social identity theory. Researchers using these models have found the donors’ identification of
5
et al., 1995). Pertinent examples of OID in the higher education setting include how individuals describe themselves, such as “I am an alumna of…” (Mael & Ashforth, 1992). This sense of connectedness begins as a student, as a positive undergraduate experience stimulates the bond to the college as an alumnus/na (Drezner & Huehls, 2014).
Mael and Ashforth (1992) stated the sense of connectedness or identification between alumni and their colleges is affected by several factors: four on the
organizational level and six on the individual level (Figure 1). First, the authors proposed that the following organizational antecedents (perceived characteristics of the college) significantly affect support: distinctiveness, prestige, intercollege competition, and intra-organizational competition. The alumnus/na’s feelings of the college as distinctive and prestigious positively affect OID. The effect of competition is complicated, as
competition between the college and another school, such as a sports rivalry, positively affects identity, while competition within the school negativity affects it.
Intra-organizational competition may be experienced when multiple departments or offices solicit the same alumnus/na for support, rather than employing an organized institutional fundraising effort (Mael & Ashforth, 1992). The authors carried out their study on a small private college because it was what Whetten (2006) describes as a holographic
organization (where organizational members share a common identity) rather than an idiographic organization (where subunit-specific identities are prominent).
Mael and Ashforth (1992) posited six factors that predict OID at the individual level: tenure, mentorship, recentness of membership, membership of similar
6
and the existence of a college mentor have a positive effect. Recentness of membership (college enrollment) also is a positive factor. Similarly, if the alumnus/na has
membership in several organizations, such as graduate schools or colleges from which they transferred, a suspected decrease in OID is suspected. Finally, satisfaction and sentimentality are positive predictors of OID. Satisfaction is defined as the alumnus/na’s overall rating of their educational experience. Sentimentality describes the feelings of fondness about their time at the institution (Mael & Ashforth, 1992). A diagram of the OID antecedent path model is provided in Figure 1.
Figure 1
Proposed Correlates of Organization Identification
Note. Adapted from “Alumni and their alma mater: A partial test of the reformulated model of organizational identification,” by F. Mael and B. E. Ashforth, 1992, Journal of Organizational Behavior, 13, p. 107. Copyright by John Wiley & Sons, Ltd.
It is worth noting that OID’s use of the concept of alumni identity has several similarities to McDearmon’s (2011) application of Stryker’s (1989) role identity theory. Both recognize the interactions between a college and an alumnus/na form a unique relationship, and the context and experiences of each alumnus/na shapes identity. However, an important difference lies in their views on motivations for support. While
7
role identity theory posits that alumni give because they believe it is what they are expected to do, OID theory argues that they give out of a sense of allegiance. One theory focuses on the cultural expectations, while the other focuses on the personal relationship. OID has drawn some criticism since its publication. Most notably, Iskhakova et al. (2017) wrote in their literature review on alumni loyalty that it is limited by its perspective. They argued that Mael and Ashforth (1992) approached the topic of alumni loyalty from the business management field, meaning they perceived colleges primarily as business entities. By framing colleges as businesses, there is a danger of only viewing alumni as valuable for economic reasons. They argued that alumni loyalty is a
complicated concept that includes aspects of material and nonmaterial, in addition to both behavioral and attitudinal. This criticism is a valuable reminder that while increasing donations is a worthwhile endeavor for colleges, the nonmaterial aspects of alumni loyalty are valuable in their own right. As an example of nonmaterial support, the authors argued that alumni are important sources of feedback on making the educational
experience better for current and future students.
An additional criticism pertains to the manner in which Mael and Ashforth (1992) grouped their variables. In their model, organizational support includes dependent variables for both donations and engagement behaviors. The model theorizes that these behaviors are similar and affected in the same ways by individual and organizational antecedents. More recent authors in the field have shown consistently that engagement and donations should be considered separately (Kroll, 2014), and that engagement activities are significant predictors of donations (Bruggink & Siddiqui, 1995; Drezner & Garvey, 2016; Hunter et al., 1999; Mosser, 1993).
8
Finally, Mael and Ashforth’s (1992) method of measuring donations is
problematic. In their study, respondents were asked to rank the school as a priority for their philanthropic contributions (1 = the school is my highest priority, 5 = do not contribute at all). There were significant problems with using this outcome measure. First, the scale did not have acceptable answer delineations. Moving from a “5” to a “4” on the survey indicates donor status, but the other differences measured priority on a Likert scale. The authors treated the differences between items the same in their
regression, leading to potential measurement error. Second, Dillon (2017) found alumni overestimated their philanthropy in surveys compared to college records. This finding casts further doubt on the Mael and Ashforth (1992) Likert scale method, and self-reporting of donations in general, as a measure of philanthropy. This belief is so widely accepted that the consultants who designed the alumni attitudinal survey in Appendix A did not include donation priority as a question and relied on the college records for donation measurements. They included a ranked priority question dealing with affiliation for survey item six, “Which of these statements best describes how XX fits in your life today? It is: One of the most important affiliations in your life today; Important to you, but other affiliations are more important in your life today; Not among the affiliations that are important in your life today.” They also included a question asking non-current donors if they gave to other organizations. It is important to note the difference between the two methods for measuring behaviors and the shortcomings associated with the Mael and Ashforth (1992) method.
In a justification for using this measure of donations, Mael and Ashforth (1992) stated total giving was not an acceptable measure due to capacity’s effect. This is an
9
important consideration supported by other studies (Mosser, 1993; Thompson, 2010). However, it is possible to control for capacity by including an income estimation, along with other variables such as class year, race, academic major, and marital status that are believed to be associated with capacity. Total giving is also relevant on a practical level, as many advancement offices use it as their primary gauge of success.
Mael and Ashforth (1992) did not use this study’s other measure, number of gifts, as an outcome variable due to “differing preferences towards lump-sum versus staggered contributions” (p. 111). While this concern is valid, recurring gifts are preferred over lump sums by many advancement practitioners because they are positive established habits rather than one-time actions. This belief fits with continuity theory, which states once an action becomes a habit, it is likely to continue (Atchley, 1989). Although the same amount of money may be given in both lump-sum and staggered scenarios, the recurring gift is viewed more favorably due to its implied association with future giving. This relevance provides further justification for measuring the number of gifts in
philanthropic research. However, Mael and Ashforth’s (1992) criticism of the measure emphasizes the clarification that the variable measures donation occurrences, regardless of size. It further indicates the formation of giving habits rather than only monetary increases, which is still a positive outcome.
Significance of the Study
The results of this study have significance for both scholars and practitioners. For scholars, the application of OID theory using SEM can illuminate the complex nature of these relationships in a visual way that aids in understanding donor motivations. By testing three methods of donations, it also showed how these relationships differ in the
10
various contexts. The analysis also applies quantitative methods to the psychological construct of alumni identity, thereby providing a powerful example of how the theory’s equation is enacted in practice. For practitioners, the results provide a visual reference for how alumni interact with their college and approach donation decisions. This
understanding aids in the crafting of solicitation methods, the structuring of long-term engagement programs, and the understanding of the beliefs that form the foundational structure of alumni identity. Most importantly, the results aid in the efforts by colleges to leverage their relationships and provide more tools to accurately predict a graduate’s inclination to donate more effectively.
Definition of Terms
The following terms have been operationalized for this study. Alumni—A group of graduates of a college or university.
Alumnus/na—A graduate of a college or university.
Alumni Attribute—Any piece of quantifiable evidence about a graduate of the school. Antecedent—A characteristic of either the individual or the college that is believed to
precede the development of OID (Mael & Ashforth, 1992).
Alumni Donors—Alumni who have given any amount of money (unrelated to tuition) since graduating from their alma mater.
Alumni Engagement—The participation of alumni in measurable activities provided by the college. Examples include event attendance, volunteering, interacting with college entities digitally, or being the parent of a student or alumnus/na. Alumni Identity—A latent construct derived from survey factors meant to represent
11
Areas of Significant Programming—A term used to denote areas in the United States where the college has regular alumni events and outreach. The areas are defined by the Core Based Statistical Area standards of the Office of Management and Budget. Cities include New York City, Washington, D.C., Boston, Chicago, Minneapolis, Denver, Colorado Springs, San Francisco, Seattle, and Los Angeles (Donovan, 2015).
Cumulative Lifetime Giving—The total amount given by alumnus/na during their lifetime.
Donor Status—The indication that an alumnus/na has given a monetary donation at any point in their post-graduate life.
Fundraising—The art and science of seeking financial support for an organization, in the case of this study, for a college or university.
Higher Education—An industry comprised of colleges and universities that grant undergraduate and graduate degrees.
Liberal Arts College—A primarily undergraduate higher education institution that focuses its primary curriculum on academic subjects distinct from professional and technical subjects. Common academic focus subjects include literature, philosophy, mathematics, and social and physical sciences.
Organizational Identity (OID) Theory—A theory that argues individuals support an organization (college) prompted by both organizational and individual
characteristics. It further states that identity is an individual’s connected state of seeing the organization’s successes and failures as one’s own (Mael & Ashforth, 1992).
12
Organizational Support—The act of giving back to the organization (college) in a meaningful way. The most common and widely studied form of support is
monetary donations, but it also can refer to acts of service such as volunteering or referring new students (Mael & Ashforth, 1992).
Predictor Variable—A quantitative variable that has a statistically significant relationship with an outcome variable.
Structural Equation Modeling (SEM)—A quantitative research method that
simultaneously models regressions and latent constructs occurring in the same theoretical system (Mosser, 1993).
Summary
Chapter I of this dissertation presented the research problem. At this crucial stage in the history of higher education, it is more important than ever for colleges to achieve financial stability by leveraging their alumni relationships into monetary donations. By focusing on behaviors that form effective engagement, the beliefs that form alumni identity, and their relationships to one another and the inclination to donate, the study provided the clearest picture yet of these complicated relationships. The study used a SEM with data derived from an alumni survey and college records. The model was composed of the constructs of alumni identity and engagement predicting donations while controlling for demographic factors. Mael and Ashforth (1992) identified that alumni connectedness drives supportive behaviors, but their investigation fell short by not fully analyzing the potential for alumni engagement to be used as a distinct predictor of OID and donations. This study showed its relevance as a predictor of giving and tests that relationship to gain a better understanding for both scholars and practitioners.
13 CHAPTER II LITERATURE REVIEW Restatement of the Problem
Colleges and universities are increasingly becoming more dependent on
fundraising efforts to achieve long-term financial stability, as government funding and tuition support can be volatile (Dillon, 2017; Horn, 2018; Lasher & Cook, 1996). The scholarly field of alumni fundraising has been addressing issues of donor motivation and engagement for decades, but much is unknown about alumni-college giving relationships. In addition, little consensus on donor motivations exists with many competing theories, offering differing views on the ways in which beliefs, experiences, and engagement behaviors relate to institutional support (Drezner & Huehls, 2014). This study seeks to contribute to the body of knowledge by utilizing a versatile statistical method, SEM, and applies it to a wealth of alumni data from a single college. The purpose of the study is to ascertain the relationship between the alumni identity, engagement behaviors, and institutional support in the form of donations. By doing so, the study tests both the strength and pattern of this model using Mael and Ashforth’s (1992) OID theory as a framework. This study also tests three measures of donations to ascertain the differences in the relationships in the various contexts.
Overview
The purpose of this literature review is to provide background on important concepts used in the research. It additionally details the history of higher education alumni engagement and fundraising. This chapter includes explanations of the major schools of thought on the reasons alumni give back to their alma maters, including a
14
focus on identity-based motivations. It also provides an overview of related prediction studies and the factors found to be significant. Finally, it discusses SEM and its application to the field.
Historical Background
According to the Council for Advancement and Support of Education (CASE) (2017a), the practice that today is known as “alumni relations” has been around in some form for over 200 years. In 1792, Yale University began the practice of organizing graduating alumni by class and entrusted class secretaries with creating biographical summaries of the graduates. The nation’s first official alumni association formed in 1821 from graduates of Williams College (CASE, 2017b). By 1913, alumni groups were common enough that 23 professional and volunteer alumni secretaries assembled in Ohio to form the Association of Alumni Secretaries, the first trade group of its kind (Keane, 1988). Yale University continued to spearhead the alumni movement by being the first to institutionalize regular reunions in 1824, the first to organize an annual giving campaign in 1890, and one of the first to publish a regular alumni magazine under the Alumni Magazines Associated organization. These practices are currently common at higher education institutions across the country (CASE, 2017a). As a barometer of the field’s evolution, CASE, the institutional advancement industry’s predominant professional association, has over 3,600 member institutions in 82 countries around the globe (CASE, 2017b).
Today, many alumni relations offices fall under the departmental purview of advancement divisions, including the site of this study. The overarching goals of
15
a college to give voluntary support (Drezner & Huehls, 2014). Thus, while constituent affinity is valuable in its own right, an explicit charge exists to turn those positive feelings into tangible support. Of these constituents, alumni comprise a large emphasis, both in terms of quantity of active constituents and the number of staff dedicated to fostering relationships. In terms of fundraising, alumni contributed 24.2% ($9.93 billion) of 2016 higher education giving amounts, second only to foundations at 30.4% ($12.45 billion) (CAE, 2017). It is worth noting that while foundations are counted separately, many of those donations are from alumni-connected organizations or alumni foundations, underscoring the importance of alumni in advancement work. All these facts demonstrate the consensus that leveraging alumni relationships more effectively has the potential to increase donation amounts (Dillon, 2017; Drezner & Huehls, 2014).
The work of connecting alumni to the college is highly subjective, but three objective industry metrics of a “successful” alumni relations program include the number of donors, the number of event attendees, and the number of alumni volunteers as a percentage of their total alumni population (Kroll, 2014). Volunteer and event attendance levels, in particular, are especially important, given that these experiences provide alumni with a more direct interaction with the college. These types of engagement activities have been shown to have a positive relationship with OID (Mael & Ashforth, 1992).
Leveraging alumni relationships more effectively has the potential to increase donation amounts (Dillon, 2017).
Relevant Theories
While Mael and Ashforth’s (1992) OID theory was deemed most appropriate for this study, it is worth noting the other mainstream theories and their criticisms. Within the
16
realm of higher education specifically, this review focuses on four additional groups of theories: altruism-based philanthropy models, exchange theories, role identity theory, and identity-based motivations.
Altruism-Based Philanthropy Models
At the heart of the discussion of donor motivation is a simple, yet controversial question. Does donor altruism exist? This question is the starting point for most of the major giving theories and, additionally, can guide practitioner solicitation. It is worth considering whether donors fundamentally give out of a selfless need to make the world better, or to get something in return, either tangible or intangible (Drezner & Huehls, 2014). For example, impact philanthropy theory draws upon a philosophical assumption that altruism exists and that donors are motivated in their gifts with the utilitarian frame to do the most good (Duncan, 2004). This theory is sometimes associated with venture philanthropy because it often refers to project-based endeavors in which expertise is also donated to build organizational capacity. Nonetheless, it inherently posits that people give based on impact rather than on self-interest (Drezner & Huehls, 2014).
Impact philanthropy is related to several other similar theories that approach the subject from slightly different perspectives. From an economics perspective, this idea is embodied in the public good model. According to Roberts (1984), altruism is “the case where the level of consumption of one individual enters the utility function of the other” (p. 137). In the public good model, by donating money individuals disregard their own self-interest (in the economic sense) to benefit someone else. Viewed through a social justice perspective, justice motivation theory draws upon the concept of altruism to suggest people donate to fix what they see as an injustice (Miller, 1977). Warren and
17
Walker (1991) posited when donors see evidence that might contradict their belief in a just world, they are motivated to restore this concept, with the strong caveat that they give only if they believe their donation will be of real and permanent help.
Finally, as an example from a psychology/sociology perspective, prosocial
behavior theory argues the motivation to give is based on the need to help others (Bentley & Nissan, 1996). The theory was developed by biologist Edward Wilson in his broader writing on the concept of sociobiology. Since prosocial behaviors can be found in a variety of animals, as well as humans, many proponents contend that this motivation is inherent and biological (Wilson, 1975). However, there is also evidence that this behavior can be nurtured and learned, and ample evidence can be found in the realm of higher education fundraising of colleges nurturing prosocial donor motivations (Drezner & Huehls, 2014). It is important to note that proponents of this theory do not associate it with altruism, but its focus on benefiting the gift recipients rather than the donor shows a strong connection to the previous theories.
Exchange Theories
In direct opposition to altruism-based theories of donor motivations, social exchange theory and its related perspectives argue that altruism is not present in the decision to give. While altruism theories include evidence of donors reporting various motivations related to impact, other researchers have been critical of this view and argue that the impact on the donor (not the recipient) is the true motivation of philanthropy (Drezner & Huehls, 2014). Blau (1964) argued that voluntary actions (including
donations) are motivated by the personal outcomes of those actions. These benefits could either be tangible or intangible. Evidence of tangible benefits in alumni fundraising could
18
involve special donor benefits such as thank you gifts, a tax benefit, or naming rights to a building. While these tangible returns are straightforward, social exchange proponents place more emphasis on the intangible donor benefits of philanthropy. These benefits are primarily internal, such as a feeling of satisfaction. However, the most powerful of these intangible benefits is the perceived elevation in social status (Blau, 1964). Kelly’s (1991) model of social exchange was used in the college fundraising context by Lasher and Cook (1996). They found it to be very present in the ways that advancement and
administration professionals engaged donors. Their methods of solicitation often involved aligning the donors’ interests with activities at the college and seeking to provide the most benefit to the donors in exchange for their gifts.
As with the altruism-based theories, similar concepts related to social exchange theory exist in other disciplines. In economics, Becker’s (1974, 1976) theory of rational utilitarianism is based on the concept that true altruism does not exist, and donors can be motivated by the peer effects of others and the need for social acclaim. Andreoni (1989) expanded on this view with warm effect theory, which argues that personal utility is a motivator, and the “warm glow” associated with donating, while intangible, is highly valuable. While these theories focus on expected benefits, the opposite may also be true. The sense of obligation, or giving because of benefits already received, is believed to be present in donor motivations, as Sugden’s (1984) reciprocity theory suggests. As an example of this theory in the higher education context, an alumnus may give to a scholarship fund because they benefited from a similar fund and have a willingness to pay back the kindness of past donors (Drezner & Huehls, 2014).
19
As a criticism of exchange models, Piliavin and Charng (1990) argued in a literature review analysis that there is sufficient evidence of altruism in action, even if it is not always present. While selflessness may not always be the donor’s motivation, it may at times be or is a component of the motivation. In regard to social exchange models that rely on social benefits, Schervish and Havens (1997) argued that while peer
motivations undoubtedly exist, altruism may be present collectively in “mutual interest” or “multi-person altruism” (p. 237). However, they also argued personal self-interest is often the primary motivator. Simmons (1991) further stated that helping acts remain admirable even when inspired by subtle self-rewards, such as the desire for one's life to matter, to improve one's self-picture, to feel happier about life and self, to relieve the distress of empathy with the victim, or to obey religious and societal norms.
Role Identity Theory
As Simmons (1991) mentioned in her analysis of altruism, the need to obey societal norms can be a powerful motivation. Evidence of this exists in higher education fundraising (Dillon, 2017). This motivation was argued in Stryker’s (1968, 1980) role identity theory, which states the social concept of self is constituted by society. Further, individual behaviors are motivated by the need to fit a perceived role, and individuals placed into socially recognizable positions will likely fulfill the social expectations of those positions (Stryker, 1968). This can be seen when individuals appointed to new positions change their behaviors to fit the implied role. Turner (2001) expanded on this theory by stating that a role is “a cluster of behaviors and attitudes that are thought to belong together” (p. 233). An individual may inhabit several (sometimes contradictory)
20
roles simultaneously, but they are primarily guided by their current context for social norms and behaviors (Stryker, 1980).
Dillon (2017) applied this concept to alumni fundraising by stating that the position of an “alumnus/na” has an inherent social expectation of donations, and alumni donate primarily because they are expected to do so by their community. A criticism around this theory is that it does not take into account the multitude of previously mentioned motivations that are undoubtedly at play. Additionally, if the role of an alumnus/na includes giving expectations, undoubtedly a logical source of those expectations should exist. While some school personnel may make efforts to set expectations early in the college relationship, little evidence can be found that this expectation is uniform and powerful enough to describe the thousands of U.S. institutions.
Finally, in Dillon’s (2017) alumni survey analysis using role identity theory, he contended that role identity is the primary alumni donation motivator, even though the analysis showed that donor status was not significant in predicting scores of alumni role identity. If role identity was such a primary motivation, it would be expected that a natural correlation would occur because all contactable alumni are often regularly solicited by colleges on at least an annual basis. While this major study did show some limitations of the theory, it advanced the field in a crucial way. The study correctly grouped donation antecedents into three distinct groups: demographics (race, gender, income, etc.); behaviors (volunteering, event attendance, social media, etc.); and beliefs (satisfaction, sentimentality, connection, etc.). By doing so, the study clarified that
self-21
concept in relation to the college is an important independent component of the giving equation.
Identity-Based Motivations
An individual’s decision to donate to his/her alma mater is a complicated and personal choice. The individual’s personal connection and a sense of allegiance to the school should be considered when examining donor motivation. Colleges have a relatively unique fundraising model compared to other nonprofit organizations that confound the traditional idea of a donor relationship. For example, alumni have already engaged in a transactional relationship with their college as students, sometimes paying large sums of money for tuition, room, board, and associated costs. This financial relationship makes the proposition of paying additional donations unusual, especially since the primary benefits of the educational experience have already been delivered. Thus, it is clear many alumni have such a strong connection to the school that they decide to give additional money.
In terms of leveraging alumni relationships for the purposes of fundraising, a wealth of qualitative research exists to shape understanding on the way in which identity serves as a motivator. Oysterman (2007) developed an identity-based motivation (IBM) model that argues actions are most natural when they align with the individual’s identity and are within context. Aaker and Akutsu (2009) expanded on IBM in philanthropy by claiming identity saliency in the realms of familial, community, or personal identity are key drivers of donations. This behavior is often seen in earmarking gifts, such as when alumni donate funds for a sport they played as a student, guaranteeing the impact will be on students similar to them.
22
While Mael and Ashforth (1992) used alumni identity in general as the main motivator, qualitative research has further explored identity-based motivations among these identity communities that exist within the college, such as those among
marginalized groups, including those based on alumni of color, gender, sexual
orientation, and ability (Drezner & Huehls, 2014). For example, Gasman (2002) found African American alumni are often ignored by advancement offices, and their
motivations in higher educational settings often are focused on racial uplift, or the need to use their gift to the college as part of a larger effort to advance their racial community. Gasman argues that this motivation is notably different than those observed in the White community, and traditional fundraising tactics are not addressing this motivation or this community.
Similarly, Drezner and Garvey (2016) found in a study of LGBTQ alumni that their giving was affected by their perception of their student experience and the campus climate of inclusiveness. They also found involvement in identity-based affinity groups for LGBTQ alumni positively affected their relationship to the college and the motivation to give. Most notably, while the alumni in their study did not directly cite identity as a giving motivation, the authors argue that it played a role unconsciously.
Significant Donation Predictors
Quantitative studies have also advanced the understanding of donation
motivations. Weerts (2007) found most academic publications on alumni donations focus on four areas: research into individual donor characteristics, fundraising practices, the external environment, and institutional characteristics. The proceeding sections focus on summarizing what other researchers have found to be significant predictors of giving,
23
particularly from the wealth of regression-based prediction studies that have identified significant predictors.
Fundraising Practices
While most of the material thus far has covered individual aspects of donor motivations, it is worth noting the college’s effect on these relationships, particularly relative to advancement practices. Harrison (1995) used a regression method to predict donor status with fundraising investment data from 18 colleges. The study used CASE financial data from three fiscal years and analyzed the relationships between giving data, fundraising expenditures, and institutional traits. The study’s general finding was not surprising. The more money schools invested in fundraising, the more money they gained in return. However, an additional finding that is particularly relevant to this study is that investment in alumni relations activities were both significant and powerful in predicting increased donor status among alumni. These findings support the argument that
engagement practices have a direct link to financial support, and investment in them by colleges should be seen as ways to bolster donations.
Economic Factors
Considerable philanthropic research has focused on how the economy factors into donation decisions, especially concerning income and the financial ability to donate. For example, the impact of taxes and deductibility has been shown to significantly affect donation decisions (Leslie & Ramey, 1988; Seltzer, 2018). Holmes (2009) also found alumni in higher tax brackets living in a state with higher charitable deductions donate more to their alma mater. However, the effect was not observed for lower income alumni,
24
showing that tax sensitivity is only a considerable factor for those more likely to itemize their deductions.
Additionally, income has continually been observed as a significant predictor of donations, as well as an accurate proxy for the concept of an individual’s ability to donate (Leslie & Ramey, 1988; Weerts & Ronca, 2009). Hernández-Murillo and Roisman (2005) even stated, "Income is by far the most important predictor of giving behavior" (p. 12). However, there is some disagreement over how the dependent variable measure of giving changes the independent variable of income’s predictive power. Van Horn (2002) found that while income is particularly significant in predicting the size of donations, it is less relevant in predicting donor status in general. This finding highlights the importance of using multiple measures of philanthropy (donor status, amount given, and number of gifts) when creating predictive models. In a similar way, marital status is believed to be a significant predictor of donations. This is possibly due to the financial stability associated with marriage, thus creating more disposable income and the ability to donate. Both Belfield and Beney (2000) and Monks (2003) found marital status to be a positive predictor for alumni donations.
Geographic Distance
Living closer to the college theoretically supports donor behaviors by providing more opportunities to visit and stay connected with the current state of the institution. Bruggink and Siddiqui (1995) and McDearman and Shirley (2009) both found that geographically closer alumni were more likely to be donors. However, both studies were conducted on public university populations, and a lack of evidence exists that this predictor is generalizable to more widespread alumni populations (Radcliffe, 2011).
25 Age
Alumni age is commonly viewed as a significant donation predictor because of its perceived relationship to income. As alumni progress in age and in their careers, they likely see their incomes increase as well (Radcliffe, 2011). Estate gifts are also quite common, and at many colleges these pledges are made by older alumni and counted as donations at the time of the pledge, although the money will not tangibly arrive until after death. That being stated, there is continued evidence that age is a significant positive predictor (Hoyt, 2004; Leslie & Ramey, 1988; Thompson, 2010). Mael and Ashforth (1992) also listed it in OID as the antecedent “recentness of membership.” However, they surprisingly listed it as a negative predictor. When they conducted their analysis, the variable was not significant on their measure of giving. They believed an individual’s connection to the organization would fade over time, which may be true, but age’s relationship to income may be transforming it into a positive predictor rather than a negative one.
Racial Identity
Race has a complicated relationship with giving. For example, Thompson (2010) found in a data mining analysis of alumni giving that White alumni gave significantly more (mean cumulative total giving) than other races to their alma mater. However, qualitative scholars such as Drezner and Huehls (2014) and Gasman (2002) suspected race is a significant predictor, not because of race’s effect on connectedness, but due to an inability of colleges to effectively solicit donations from non-White donors. Similarly, the factors may also be affected by the history of racial wealth inequality in the United
26
States, as well as negative student experiences. As such, race may be a useful control variable in prediction models.
Degree Type
Degree type is seen as a predictor because of its effect on income. For example, students in professional and technically focused degree programs consistently earn higher starting salaries (National Association of Colleges and Employers [NACE], 2018). Similarly, Thompson (2010) found that attaining a Bachelor of Science degree, as well as a joint Arts and Science degree, were significant positive predictors of donations.
Blumenfeld and Sartain (1974) also found being an Economics major was a significant predictor of giving. Degree type effectively serves as another measure of ability to give and, thus, can be a useful control variable.
Student Experience
Perhaps the most important characteristics of Mael and Ashforth’s (1992) OID theory deals with how alumni view their student experience, and the proceeding research has justified these claims (Drezner & Huehls, 2014). Gaier (2005) found highly
compelling evidence of the importance of satisfaction from a large-scale survey of graduates from a large public university. Not only was there a significant and powerful connection between satisfaction and giving, but also with student involvement. The study showed alumni who participated in a formal activity as students were 87% more likely to be donors. These findings, along with Dillon’s (2017) analysis show the importance of including survey data with indicators for student experience and satisfaction.
27 Alumni Engagement
As mentioned previously, Mael and Ashforth’s (1992) OID theory views alumni engagement behaviors, such as event attendance and volunteering, as dependent variables representing organizational support. However overwhelming evidence exists for these behaviors to be seen as significant predictors of OID and giving (Drezner & Huehls, 2014; Hunter et al., 1999). Wunnava and Lauze (2001) found in a study of Middlebury College’s giving records over 23 years that significant predictors of alumni giving were as follows: volunteering for the college, residence in states with alumni chapters,
existence of relatives who are alumni. Similarly, Thompson (2010) found in a large-scale donation records analysis of eight colleges that formal participation in alumni
associations and more informal information sharing were significant predictors of alumni giving. While Mael and Ashforth (1992) did find evidence that these behaviors may be effected by OID and the supporting antecedents, the use of alumni engagement as a predictor shows the value of using them in a dual role of being affected by, but also affecting, OID.
Structural Equation Modeling (SEM)
While most of the previous studies have used straightforward regression
techniques, it is worth noting SEM’s particular relevance to this field of study. Mosser’s (1993) study on predicting alumni giving behavior is the only available example of SEM modeling involving alumni giving. He utilized a massive survey dataset of 110,000 alumni at the University of Michigan. While the study was used effectively to predict key indicators of alumni giving (willingness, capacity, and academic integration), the study utilized no variables related to engagement or latent constructs related to identity. The
28
author also concluded that while the study was helpful in determining capacity’s
mediating relationship to donor motivations, there are many other factors that he was not able to take into account. Since then, there has been a continued literature gap and need for SEM use in alumni fundraising. Heckman and Guskey (1998) even called for its use in future research in their study using a path analysis. They applied the economics-based discretionary collaborative behavior theory to an alumni survey. The path analysis was helpful in visualizing the relationships, and they provided evidence for the importance of alumni-college giving relationships, while framing it instead as a customer-business relationship. Most notably, rather than using donations as the primary dependent variable, they used a scale variable that counted up to 20 positive behaviors, weighting donor status (occurrence of any gift) equally with acts such as hiring an intern or volunteering with admissions.
Summary
The wealth of relevant donation literature indicates the decision to donate money is complicated and personal in nature. Additionally, it should be clear that the
motivations of alumni donors are inherently unique in the world of philanthropy. While universal notions of achieving impact and social exchange have been observed, the identifications of alumni with their alma mater are undoubtedly present and significant. This sense of identification connects alumni to the college to the extent that they experience the college’s successes and failures as their own. The literature also has shown this concept is an independent component of the donor motivation process. The other two key groups in the process include behaviors in the form of engagement
29
activities and demographics in the form of significant predictors inherent to the individual, such as race, income, and location.
While the preceding studies are helpful in providing findings on donor
motivations, there is still a need for a more in-depth understanding of the relationship between engagement behaviors, the concept of alumni identity, and the outcome of institutional support through donations. As donor motivations are complicated and primarily involve finding measurable predictors of future behavior, straightforward regression models have been the primary method for quantitative studies (McDearmon, 2011; Thompson, 2010; Wunnava & Lauze, 2001). The consistent problem with the application of this methodology is that these studies usually focused solely on either attitudes (satisfaction, sentimentality, and connection) or behaviors (events, volunteering, and student involvement). Both categories are associated with donor status, but a unique literature gap exists in the need to analyze and map the relationship between all three (engagement, identity, and donations) in one model while controlling for demographics. Specifically, the use of SEM methodologies is needed to utilize the application of regression correlations in a complex model that considers several sources of data on the same sample of alumni. Doing so furthers the understanding of this complex relationship using Mael and Ashforth’s (1992) OID theory, with the notable change of including engagement behaviors as both forms of support and donation predictors.
30 CHAPTER III METHODOLOGY Restatement of the Purpose
This chapter presents the methodology and procedures of the research. A description of the survey instrument, the other data, the sample, and the methods of analysis is included. This study was designed to map and analyze the alumni-college giving relationship. Specifically, the following research questions guided the study:
1. What is the nature of the relationships between engagement behaviors, alumni identity, and donations?
2. Do the relationships follow Mael and Ashforth’s (1992) OID framework? Research Design
This study employed a quantitative analysis that involved SEM to analyze data from a self-report survey of college alumni and the college’s alumni database. The SEM methodology is ideal because it is designed to test the true nature of the alumni-college giving relationship. The model used is what Byrne (2010) calls a measurement SEM, as opposed to a structural model, because it involves both observed and unobserved
constructs. The model is also recursive because there are no feedback loops. The research was conducted in two overarching steps. The first was to analyze the survey data to find the best numerical representation of alumni identity. The same procedure was carried out on the college’s records to find suitable variables to form the latent construct of
engagement. Reliability analysis procedures were performed to ensure the individual items suspected to form the latent variables of alumni identity and engagement were acceptable for the model. The next step involved the use of SEM to map the relationships
31
between the variables. The model provided an effective way to visualize how the three concepts of engagement, identity, and donations relate to one another, while controlling for a variety of factors. The variables were categorized as multiple factors in four groups: donations, alumni identity, engagement behaviors, and demographics/control variables.
Research Setting
The setting was a small liberal arts college in the Western United States. The school is selective (15% acceptance rate in 2018) with a student body of approximately 2,200 and an estimated living alumni population of approximately 30,000. The college has a distinctive academic curriculum in which students take classes one at a time for three and a half weeks, for a total of eight classes per academic year. The college focuses primarily on undergraduate liberal arts education, with only one master’s program in teaching. The college offers 41 undergraduate majors, 27 traditional minors, and 31 thematic minors that emphasize interdisciplinary studies. The top three most popular majors are Economics, Biology, and Sociology. The majority of graduates go on to pursue advanced degrees. Over half (53%) of alumni were either enrolled or completed an advanced degree by five years after graduation.
The college is the only selective, private liberal arts college in its immediate region, and it is a member of the Associated Colleges of the Midwest (ACM). Its campus lies in an urban setting. The college is in a relatively sound financial situation with an endowment of approximately $765 million ($363,000 per student). The campus reflects this state with several new or newly renovated buildings, including a large recreation center, performing arts hall, fine arts museum, and a state-of-the-art library, which is the largest carbon neutral academic library in the United States. The campus is primarily
32
residential, with 78% of students living on campus. The average class size is 15, with a student-faculty-ratio of 10:1. The college has a strong first-year retention rate of 96%, with a four-year graduation rate of 82%.
The student body consists of 64% White students. The races and ethnic groups self-identified by students of color are 15% Asian, 9% Hispanic, 5% African American, 2% Native American, and <1% Native Hawaiian/Pacific Islander. These statistics include some double counting due to those with two or more races identified. International students make up 8% of the population. The surrounding city is 78% White according to the 2010 census. Its international students come from over 50 countries, with the largest population from China (110 students in 2018). The tuition cost for the 2019-2020 school year is $57,612, with a total estimated cost of $74,760 for residential students. The tuition cost has risen by at least 3% every year since 2010. While the listed cost of attendance is high, the average net price is $28,390 per year due to the various forms of aid. Half of current students receive financial aid, and 12% receive Pell grants. The college’s sizable endowment ensures that even students paying the full tuition price are partially
subsidized and receive services valued at more than their cost. The largest group of students is from the East Coast, with 28%.
The college is in the midst of a large capital campaign, providing additional relevance for a study examining the relationships between beliefs and behaviors of alumni. The school has two Division I sports programs, men’s hockey and women’s soccer, and disbanded its Division III football program in 2008. The college holds regular alumni events across the country in major cities and maintains several alumni volunteer programs. The college currently includes 55 full-time employees in its advancement
33
division. The alumni office utilizes a number of volunteer/engagement opportunities, including reunion committees, advisory boards, career center volunteering, nationwide events, and an online mentoring community.
The college population is ideal for an examination of OID because it objectively rates highly in all four organizational antecedents in OID theory: distinctiveness, prestige, intercollege competition, and low intra-organizational competition. It is distinctive due to its curriculum and unique geographic location. It is prestigious in the sense of having highly selective admission practices. While the college is not known primarily for athletics, it has a notable Division I sports rivalry contributing to intercollege competition. The college’s small size also limits the potential for intra-college competition. Students have more opportunities to interact with others outside their academic major, and the vast majority of official reunions are organized around class cohorts instead of disciplines. The college is also ideal because it matches the type of college (small private institution) Mael and Ashforth (1992) used in their study. Mael and Ashforth felt that the use of such a college was ideal for research because it was what Whetten (2006) describes as a holographic organization (where organizational members share a common identity) rather than an idiographic organization (where subunit-specific identities are prominent).
Population and Sample
The population for the study was derived from those who participated in a college survey in the fall of 2018. The sample size for the survey was 1,922 (1,777 completed it and 145 answered some of the questions). The survey was conducted by a third party, eAdvancement, a consortium of independent consultants in alumni relations,
34
communications, and fundraising (eAdvancement, 2018). The survey was emailed to a representative sample of 7,823 alumni, resulting in a 25% response rate. The study was commissioned by the college’s advancement division, and permission was granted from both the college and the survey authors to use it for this research. In addition to the report, the raw data with identification numbers were available. This level of data allowed for the matching of survey responses to engagement and giving behaviors. The list of survey questions used in the study and the percentages for each answer are provided in Appendix A. The demographic breakdown is provided in Table 1, with comparisons to the demographics of the entire alumni population.
Table 1
Location, Gender, Graduation Years, and Donor Status of the Survey Respondents Actual Population
%
Survey Respondents %
Location
Metro areas near the college 18% 23%
Areas of significant programming 22% 29%
Everywhere else 60% 48% Gender Female 50% 52% Male 50% 48% Graduation Years 2010-2017 14% 17% 2000-2009 17% 21% 1990-1999 16% 20% 1980-1989 15% 19% 1960-1979 21% 24%
Donors (prior three fiscal years)
Donors 31% 36%
35
Instrumentation Alumni Survey
The survey consisted of 28 question groupings with additional follow-up questions depending on the answers. The questions were derived to measure a range of alumni opinions about the college, including communications, activities, and donor motivations. Several questions were designed to measure concepts related to alumni identity, those items were composed of three components: student experience satisfaction, current sentiments on the college, and positionality as a member of the community. This study used 12 of these questions (Appendix A). The majority of the questions used a Likert scale, including those believed to be associated with alumni identity. For example, alumni rated their overall feelings about the college on a scale of 1 to 5, with 1 being very dissatisfied and 5 being very satisfied. In another example, the alumni were asked how much pride they have in their college affiliation. That particular question used a four-item scale with 1 being not at all and 4 being a great deal.
The survey consultants provided the college with net promoter scores using the question, “If you were asked, how likely would you be to recommend the college to a prospective student of your own interests and background?” using a scale of 0 to 10, 0 being not at all and 10 being highly likely. The net promoter score concept was
developed by Fred Reichheld, and it categorizes Likert scale questions into three groups: promoters, passives, and detractors. Promoters are loyal and enthusiastic supporters, passives are satisfied but unenthusiastic, and detractors are unhappy constituents who could cause harm to the brand (Reichheld & Markey, 2011). The groupings for the net promoter question were 9-10 (promoters), 7-8 (passives), and 0-6 (detractors). The
